That's why I was surprised about the option to inject a bean directly... it
almost sounds like going through the NMR would be worthless in that
respect. So it's just another option if you didn't care to take advantage
of the functionality the NMR provides.
On 8/29/07, Bruce Snyder <[EMAIL PROTECTED]> wrote:
>
> On 8/29/07, Ryan Moquin <[EMAIL PROTECTED]> wrote:
> > While I'm thinking about it, is there any reason for going through the
> NMR
> > vs. just using the bean directly? Is it only to be loosely coupled or
> are
> > there any other advantages?
>
> The NMR mediates communication so that the location of each service
> doesn't matter. Use of the NMR is done via a flow and the three main
> flows provided by ServiceMix (i.e., SEDA, JMS and JCA) each provide a
> certain level of threading and buffering for messages via
> java.util.concurrent queues.
>
> Consider a situation where you have multiple instances of ServiceMix
> running. When networking the containers together, the NMR allows
> services running in one container to communicate with services running
> in another container very easily. If you make those two services talk
> directly to one another without using the NMR, then you are completely
> circumventing the inherent benefits provided by the NMR. In a way,
> this would be no better than running web services in stand alone web
> containers and making them talk directly to one another.
>
> Bruce
> --
> perl -e 'print
> unpack("u30","D0G)[EMAIL PROTECTED]&5R\"F)R=6-E+G-N>61E<D\!G;6%I;\"YC;VT*"
> );'
>
> Apache ActiveMQ - http://activemq.org/
> Apache ServiceMix - http://servicemix.org/
> Apache Geronimo - http://geronimo.apache.org/
> Castor - http://castor.org/
>
